Some people believe it is art, while others say no way. Artificial image generation is a relatively new and fascinating technological development in the market. But it raises a lot of questions regarding artistic, creative, ethical and legal (i.e. copyright) aspects.
By definition, “art” stems from the Latin word “ars” and implies things like acquired skill, craftsmanship, and artistic achievement. It describes the expression or application of human creative skill and imagination. The visual end result is meant to be appreciated for its beauty and emotional power. This can be achieved, for example, by the artist sharing their personal story through their art. Lastly, it also involves components like critical thinking and problem solving capabilities.
Can these factors be found in AI-generated artwork, too? Rather not. At least not to the same extent as in human-created art.
As artificial intelligence seems to become an overwhelming force in the creative industry, it raises some difficult questions and a problematic outlook on its future role in the art world.
Is AI art real art? Should it have the same value as traditional or digital art that was fully created by a human? How much human processing is needed for AI art to still be considered 'real art'?
Related: Can Fine Art Be Digital? [Addressing The Digital Art Debate]
I’ve long hesitated to touch upon the AI Art vs Human Art debate on my blog because I know it’s a very controversial topic that got heated-up even more after an AI-generated work recently won first place at a State fine arts competition, and “artists are pissed” (quote ☺︎).
But is important that we – especially as artists – talk about it. So here we go! With this blog post, I am actually starting a series of posts addressing the AI Art vs Human Art controversy.
My intention is neither to glorify AI art nor to demonize it. I rather want to launch a discussion and invite artists to share what they think of AI art.
I’ve already experimented with AI-generated images myself and used some of those images as background in one of my surreal creations. But more on that later on in this post.
In this blog post, I want to address the question whether AI art can be considered real art, share my opinions, and put the topic up for discussion in the comments.
So is AI-generated art really art? Let’s unpack it!
Can AI create original art?
First off, what exactly is AI-generated art? In a nutshell, AI-generated art refers to images that have been created with the help of artificial intelligence. There are different platforms that allow you to generate that type of images. Midjourney, Stable Diffusion and DALL-E are among the most popular ones.
Such image generation systems produce visuals based on the parameters that the user provides in form of natural language. Once the user has typed in their so-called prompts, an algorithm will generate an image that corresponds to the user input.
Usually, the AI systems offer you a set of images to choose from. You can pick the image you like best, click again to have the AI make further variations or scale it up to a high-resolution image for you to download.
The prompts don’t have to be complicated. They can be short and descriptive. Oftentimes, simple prompts produce better results than long and complex text input. In fact, you can create impressive and highly intricate images in just a couple of seconds. On the flip side, you can as easily create ‘artworks’ that imitate the style of a particular artist or art movement.
For example, a prompt like “piano in a starry night vincent van gogh” will make the algorithm come up with images that depict the music instrument against a night sky pretty much the way the Dutch post-impressionist would have painted it. Likewise, typing in something like “abstract art in the style of picasso” will produce results that are the spitting image of what Picasso would have created.
So is AI art a threat to artists? Here are some examples that illustrate how easy it is to steal an artist's style using AI tools:
Three examples of AI-generated art that mimic the style of famous artists. Left image in the style of Van Gogh’s painting ‘The Starry Night’ (image via PZK3759 on YouTube) and two AI pieces in Picasso’s style on the right (images via Colosssyan6874 on YouTube).
Of course, it is the user’s responsibility to not use AI image generators in inappropriate and unethical ways. But should it be so easy to create carbon copies in the style of famous artists? After all, a human painter can also counterfeit another artist’s work and sell it as the original. But undeniably, they require much more time, effort, skill, and practice to do so! (More about that topic in my third post: Is AI Art Stealing From Artists?)
So much for the creation of AI images. What are you thinking? Is it fair to consider AI art to be real art?
Are you an artist if you use AI?
In August of 2022, an AI-generated artwork won first place in a fine art competition in Colorado.
Jason Allen, president of a gaming company called Incarnate Games, submitted a canvas print of “Théâtre D’Opéra Spatial” (French for Spatial Opera Theater), a piece he had created using Midjourney. The contest’s definition of digital art was broad enough to make AI work eligible. And to what Allen says, he clearly labeled it as a piece created via Midjourney. Still, artists were outraged and started to raise ethical concerns.
Allen explained that he had spent hours over hours figuring out the right prompts that eventually enabled him to create this piece. He added that he processed the work in Photoshop and then finished it in Gigapixel.
Allen is aware of the controversy his submission caused but is convinced that the art world will eventually recognize AI art as its own category.
His reasoning behind why AI art should be considered real art?
Well, Allen suggests that the difficulty of creating an AI piece through a complicated series of prompts may be comparable to a painter making “their art while hanging upside-down and being whipped while painting.” Still, the art of that painter is evaluated the same way as the work of a painter who creates a similar piece ‘normally’.
In other words, Allen believes that AI-generated images are not less valuable than man-made art just because AI art is easier to create.
How do people argue that AI art is art?
As we already said above, some people argue that AI art is not art because it is created by a machine. On the other hand, there are those who believe that AI art is no less valuable than any other form of traditional or digital art.
Here are the most common arguments I came across that people bring forward to establish that AI art is art:
• AI art requires human involvement to create it.
Before any artificial intelligence can generate an image, it needs a human to provide the parameters that describe the desired result.
• AI art is more authentic because it doesn’t try to conform.
When human artists create art, they often try to comply with conventions in order to not be the odd one out. Humans are also more likely to make art to please people.
• Creating AI art takes time and practice just like other forms of art.
Finding the right prompts to create what you have in mind takes time and sometimes hundreds of trials. AI image generators also come with a learning curve.
• AI art generators need time to collect data and learn themselves.
Text-to-image generators are in the middle of a learning process themselves. They need to collect data and rely on human feedback for their algorithms to figure out which generated images are good and bad.
• AI is just another tool to create art.
AI is a tool just like Photoshop or any other digital art software to provide you with the technological means to create visuals.
• Selecting the best AI image requires the trained eye of an artist.
While anyone can enter prompts, only the experienced artist is able to see which images have the potential to make a great artwork after further modifications and processing.
In conclusion, can AI art be considered real art?
Before sharing any of my personal opinions about whether AI art can be considered real art, I want you to know that I already tried out AI-generated images myself, using Midjourney. It’s not that I’m talking about something I have absolutely no idea about… ☺︎
So in which of my pieces did I use some AI art?
Scared Bunny is the third piece in a series of three (so far) that I created in 2022. The other two pieces are I survive (Fingers crossed) from 2017 and Cat & Mouse from 2014. (You can learn more about the meaning and making of Cat & Mouse here.)
Now to my AI experiences. When AI started to become such a big hype, I got curious and decided to give it a try. On the free Midjourney trial, I played around with a prompt like “green bamboo forest with big colorful flowers,” and got back a set of four pictures that looked pretty decent. I chose two of the images and actually merged them together to form the background that you can see in my above artwork.
Everything else is created by yours truly using Photoshop and a drawing tablet; the face is my very own as always in my Surreal Stories.
(Read more about my experiences with Midjourney in my second post: Why Artists Question That AI Art Is Ethical.)
So how did I feel about creating images with artificial intelligence as an artist?
So my thought process went something like that: “Wow, these pics look quite pretty. And they were created so faaaaast! Is it really so easy?!”
After all, my first attempt already brought about a range of images that are pretty complex and aesthetically appealing. Without any previous practice, without any trial and error. It just happened like that.
If I were to create the same background in Photoshop, it would have taken me hours and hours of work, along with some slips and adjustments. No way to create such a highly detailed visual in just a couple of seconds.
…and that would also be my first point of criticism when it comes to arguing whether AI art can be considered real art:
• Creating AI art is child's play.
AI images are so super-easy to create that every child can do it without any prior knowledge whatsoever. It’s fascinating. Simple prompts can generate really complex and intricate images in no time.
In other words, anyone can create AI art, regardless of their artistic ability. You don’t even need any physical materials like a brush or a graphics tablet.
Now, I already see AI enthusiasts wanting to jump down on my throat and screaming that AI prompts need to be worked out and refined, too. Sometimes even hundreds of times before you get the image you want. (Not to forget that the AI is highly addictive! It sucks you in and compels you to try more and more images. I noticed that myself when I tried out Midjourney. This addictive nature is also one of the reasons why artists are against AI art.)
But honestly, how much time do you need to learn and apply those commands as compared to how much time you need to learn and master programs like Photoshop?
Moreover, all the prompts a user enters to create a particular image are visible to all other users. That means any user can copy those prompts, modify them a little bit, and create a similar piece of art. They can even use the same prompts. Still, the AI will never give you exactly the same image, because its algorithms are experimental and random (for copyright purposes).
• AI art is not original.
This is perhaps the biggest and most common objection that people raise against AI-generated art. It cannot be original.
All artificial text-to-image generators can do is crawl through millions of data sources and patch together pre-existing images to create an output that comes as close as possible to the described subject, concept, artistic style, and so on.
If no human had created those images in the first place, AI now wouldn’t have anything to draw from. This also means that AI will take from its own images to create more art in the future. Huh… where are we going?
Now, you could say, ‘but Angie, you are also using pre-existing images to make your photo manipulations’. Yep, but I am only using stock images that are made for that purpose. Do you think the AI does so, too? (Hint: No, it does not. More about that in Why Artists Question That AI Is Ethical.)
Plus, I always give my works a personal note by using my own stuff, such as photos of myself and of other things, with a lot of additional painting and editing.
I would say that my art always includes something personal, in many cases I tell a personal story through my art – which brings us to the next point…
• AI art is not personal.
This is one of the major differences between AI art and human art. For the human artist, the inspiration to create art comes from their experiences, memories, emotions, cultural background, etc. They use their creativity and imagination to convey a specific concept through a visual piece.
The creative process triggers complex organic-neural networks because it requires the brain, hands and sensory organs to collaborate harmoniously to achieve the end product.
Artificial intelligence, on the other hand, is triggered by commands only, not by emotions or the desire to connect with other humans through the art it creates.
I can already think of the counter-argument that AI artists will bring up here. What if the human enters their prompts with feeling and the intent of conveying a personal story through the AI art they create? Well, yes ok. But then they are bound by the results the AI spits out, because you are never going to get the exact image you have in your mind.
Human art is much more genuine and straightforward in that regard. I would also raise the following question here: is making AI art actually as satisfying to the artist as creating the art that is dear to their heart the ‘normal’ way?
Let me wrap this post up here. I also don’t want it to get too long so that no one wants to take the time to read it…
AI art is still pretty much at its beginnings and we don’t know yet where it’s going. It’s an extremely multi-facetted topic, and in this blog post I’ve tried to look at it from different angles.
So in the end, what's my take on AI art? Well, in my opinion AI art can not be considered real art because it lacks the essential components that define art as such.
In particular, it is way too easy to make to be considered a craft. Anyone can generate AI images right off the bat, without having any idea about the actual human creative process. In fact, AI deprives you of putting the time, effort, and hard work into acquiring the skills necessary to create the art you want.
I intentionally say ‘deprive‘ because I consider the development process of your artistic skills to be an important asset that also helps you in your personal growth. I speak more about that in my post on what artists think of AI art.
Then, there are the copyright and ethical issues. Since artificial image generators collect their data from millions of sources on the web, they inevitably ‘steal’ from every image they use to render a user request. This concerns both the ‘small’ and the great artists in history, and anyone who ever posted a picture online.
The matter gets even worse when people prompt the AI to imitate the style of a particular artist. Artists put in years of hard work to develop a unique art style, and artificial intelligence may rip it off within a split second.
AI art can help artists if it serves as inspiration, reference, or at the most as a side element in a visual composition. But it should never fully replace the actual ‘human’ creative process.
What you are thinking about AI art? Do you consider AI art as real art? Please leave comments and questions below or reach out to me directly. I’d love to hear from you! Make sure to sign up for my email list below so you don’t miss out on new blog posts and other cool stuff. ♥
For more about the topic, check out more posts from my AI Art vs Human Art series:
AI art is just a gimmick. Real art comes from human emotion and experience. Machines can’t truly understand or express those deep, personal feelings that make art meaningful. Let’s not dilute the essence of true creativity with artificial intelligence.
Hi Damian, I agree that AI-generated images lack emotion and experience. Every time I see an AI image, I’m wondering who is actually the one creative: the human entering the prompts or the machine making something out of them?!
I appreciate the exploration of AI in art, but I think it’s important to distinguish between art created by humans and art generated by algorithms. While AI can produce interesting visuals, the emotional depth and intention behind human-created art are irreplaceable.
Hi Jonathan, I actually hold the same view. Art requires the human element in order to be considered art. Even if the promptor appoaches the generation with emotion and intention, the AI still creates *without* feelings and emotions…